From the 2014 Strata + Hadoop World conference in New York City, an interview with the Vice-President of Strategy at Silicon Valley Data Science about the Data Lake dream, the avatars of the space, and the parallel between Hadoop and Linux.
Data Strategy
Deep Learning: Intelligence from Big Data
Deep Learning begins to derive significant value from Big Data. It has already radically improved the computer’s ability to recognize speech and identify objects in images, two fundamental hallmarks of human intelligence.
Moderator
Steve Jurvetson, Partner, DFJ Ventures
Panelists
Adam Berenzweig, Co-founder and CTO, Clarifai
Naveen Rao, Co-founder and CEO, Nervana Systems
Elliot Turner, Founder and CEO, AlchemyAPI
Ilya Sutskever, Research Scientist, Google Brain
New York Times – Future Data Strategy
This internal New York Times innovation report reveals the future data strategy of the newspaper.
Traditional newspapers have to close. Even if these newspapers are producing great journalism they are failing in the art and science of getting their journalism to readers. They haven’t done enough to crack that code in the digital era.
The urgency of a digital first strategy is only growing because digital media is getting more crowded, better funded and far more innovative.
The New York Times envisions a strategy of growth, speed, agility and unlocking the power of data.
The focus of the digital first strategy of the New York Times is about getting more people to read more of the newspapers’s journalism.
This should also be achieved by using a “Data Strategy”.
The base for the strategy is the awareness to invest more in the unglamorous but essential work of tagging and structuring data.
The Times is hamstrung in theirs ability to allow readers to follow developing stories, discover nearby restaurants that the Times has reviewed or even has its photos show up on search engines.
Metadata is the key to understand and find the content in the Times archive. Tags that tie together articles, photos, and videos and geographic coordinates will structure the data to replace the current Times index. This should allow the ability to break up each article into atoms of news, such as facts, quotes, and statistics. To enable editors to quickly surface relevant content and context during breaking news.
The next step is to improve content personalization. Through the us of better algorithm based on the help of newsroom’s editors and analysis of readers reading patterns being planned by Design for NYT5 and the iPhone app.
From there the opportunity arise to track realtime data from users behavior on all media channels. The smart use of reader’s data is seen as essential to the future success to engage more readers . The Guardian, for example recently launched its “Known Strategy” with the newsroom leading the effort to improve how it collects, archives and uses reader data.
Inside the Times there is already a analytics groups which uses data to learn about their readers, changing habits as well as the effectiveness of their advertising and marketing. They also gather direct feedback from their readers about what they want from their apps and websites. This group translates those needs for Product and Design. But, data analytics should also be used for the newsroom to better understand reader behavior, adjust to trends and drive traffic to the Times journalism. Analytics skills are needed in many parts of the newsroom, including for top-level strategy as well as desk-level decision-making.
Looking on the New York Times Data Strategy from a data company’s perspective: The approach is still very people centric. There is a lac of understanding of the technology already available, to go beyond the limited vision shown in this report. If the New York Times is just focusing on the strategy displayed in this report they will even lose more ground to their competition. The focus should be “context”. Putting more than not just their archives and realtime data in relationships but furthermore the whole news ecosystem.
The New York Times is producing content since 1851, it is a “Data Company”. Therefore its should act as such
The Times want to hire analytics experts to work with news, platform, and product editors, newsroom strategists and the people trying to grow the Times audience.
Or hire Data Without Limits (DataWL)
NYT Innovation Report 2014